Share page with AddThis

ALPHA: Clouded returns

There is a tendency to see alpha as just another word for active management. But Angele Spiteri Paris finds alpha is about more than just beating a benchmark return ...

clouded_returns.jpgInvestment chiefs at Sweden’s large national AP7 pension scheme sound a little jaded by claims from hedge funds that they can provide alpha returns. Exposure to hedge funds has been scaled down exactly for this reason. 

“We have realised that moving from funds of funds into hedge fund replication products can help us get attractive beta exposure without fooling ourselves that it is actually alpha,” says Richard Gröttheim, CIO at the fund.

When an investment firm claims to be able to produce alpha, it is saying it can gain greater returns on a risk-adjusted basis, or beat a benchmark return without veering away from the risk inherent in that benchmark. It implies the manager has a process, or a piece of information, that no-one else has that affects an asset price.

But just as capturing alpha is a difficult task, for the client or adviser whose job it is to discern how much of a return is down to alpha, cutting through all the hubris that can cloud manager returns is also a fine art.

A survey conducted by Edhec Risk & Asset Management Research reveals that two-thirds of respondents use absolute performance in a peer group to analyse the alpha they generate. But this is not an accurate means of measuring alpha because it is difficult to prove that all managers in that peer group have like-for-like investment styles, the Edhec European Investment Practices Survey 2008 states.

Felix Goltz, senior research engineer at Edhec, cites the example of small-cap Eurozone managers. “All the managers within this peer group invest in small-cap Eurozone stocks, but some may have a value tilt, while others may have a growth tilt.”

These style biases usually contribute to a large portion of returns.

Reza Vishkai, head of alternatives at Insight Investment, agrees, saying: “It has become a lot easier for people to claim, ‘We’ve outperformed by so much and therefore our alpha is this’. However, in any such statement you are missing information on the risk profile of the investment.”

Aside from the construction of peer groups, ranking manager skill according to their track record is also a doubtful practice.

Rick di Mascio, CEO of Inalytics, says: “All the track record does is tell you whether you’ve won or lost. It says nothing about how you played the game.”

Andrew Parry, CEO, Sourcecap International, says that the longer a track record is, the more it can reveal about the nature of an investment process, because there is a limited number of times a manager can strike gold over a long period of time without there being a scrap of skill involved.Parry says: “Get ten colleagues to toss a coin five times in a row and have heads represent annual outperformance and tails to represent underperformance. Then see how many get five heads in a row to represent five consecutive years of alpha generation. It’s entirely down to luck!”

Alpha blaggers
There are ways for managers to claim alpha when in fact no alpha is present. Nick Sykes, worldwide partner, Mercer Investment Consulting, says: “In a number of asset classes you can generate outperformance through allocations to assets riskier than the benchmark and look back and claim outperformance, attributing it to manager skill.”

Sykes says that adding riskier assets to a fund should be done by fiduciaries in line with risk budgets, not by asset managers seeking outperformance.

But non-benchmark assets can be included in a portfolio tactically and this can actually reveal manager skill.

Parry of Sourcecap comments: “For the skilled active manager willing to take bets against a benchmark that is a blessing, but for those that cling to the benchmark it can be a disaster.
“If they are skilled they are heroes, if they aren't then they are doomed muppets.”

As well as fluking returns with riskier assets, there is also the possibility of managers just packaging good beta returns as alpha.

But generating additional returns through beta does involve manager skill, as John Hastings, partner at Hymans Robertson, recognises. “If you can find a manager who knows what he’s doing and is an alpha packager of exotic beta then that is a perfectly acceptable form of active management,” he says.

However, the issue lies in the marketing of these returns as alpha rather than beta, which is cheaper to obtain. But Hastings believes investors should not be turned off making an allocation to exotic beta that is labelled as alpha as long as there is skill to be had.

“Essentially, if they can package something for you that you cannot obtain anywhere else, why shouldn’t you pay them if you want to get it?” he says.

Cruder still
The Edhec survey shows that some managers also choose to measure their ‘alpha’ by looking at their return compared to the market index.

“This is a very crude way of defining risk,” Goltz, of Edhec, says. “Taking this value means that no differing investment style or credit risk is accounted for.”

Mark Lyttleton, a fund manager at BlackRock, says: “People will market funds in many different ways and it’s up to investors to really get to the bottom of what is really going on within each fund. They have to be careful before they invest and understand that people can promise anything but it is critical to understand the investment process and in what conditions it will perform.”

But a more in-depth understanding of the nuances and layers of risk managers expose themselves to can be brought to light through the adoption of customised benchmarks – although these have been seen to be a contentious topic.

Goltz of Edhec explains that the correct definition of a benchmark can adequately track all the risk a manager is taking and hence lead to a more detailed picture of the alpha generated. A holdings-based style analysis allows for a complete breakdown of the risk exposure, thus revealing the source of returns. A returns-based style analysis is cheaper and more practical than holdings-based analysis. However, it does not give the clear and precise picture that running a fine-toothed comb through the holdings does.

Hastings of Hymans agrees that holdings-based style analysis is ideal if one were to have the
time and resource to carry it out every time. “Given a choice I would rather use holdings-based style analysis. Returns-based analysis is a look back measure and past performance is useless – you can’t buy past performance.”

© 2008 funds europe